Developing two-wheeled inverted pendulum robot with vision based navigation and obstacle avoidance algorithm
The main purpose of this thesis project is to develop computer vision based navigation for the two-wheeled, inverted pendulum robot, that will enable the robot to detect corridor, recognize the obstacle ahead, and avoid the obstacle. The navigation system, however, is best implemented in an indoor environment. Vanishing point, which is produced by the corridor detection algorithm is used as a guidance point for the robot navigation. Ping sensor and camera based vision will support each other to provide a reliable data for the Obstacle avoidance algorithm. The balancing system is also modified in order to be compatible with the navigation algorithm. The secondary purpose of the thesis is also to establish continuous communication between mini-PC, master microcontroller, and slave microcontroller. This continuous communication, plays an important part in the two wheeled, inverted pendulum robot system, in order for the system to be able to pass the right data at the right times without interfering with its balancing system. Some experiment also conducted on the system in order for the robot to be able to use different IMU.
B01804 | (Rack Thesis) | Available |
No other version available